I get that this may be unappealing to many, but is there a way to take a (finite) quasigroup (that is not a group, but is otherwise as nice as you like) and build a (finite or at worst countable) group that contains the original quasigroup (at least the elements and products).
I've tried a few things, but it seems that there's no reasonable way to "repair" the quasigroup. I was thinking of 2-groups and how they are basically crossed modules of groups, so you have elements and functions and they intermingle, but it is an operation that is kind of "closed" under groups. Similarly you could t…